Re: [squid-users] Reverse Proxy with Multiple Backend Web Servers

On Wed, 9 Sep 2009 13:37:04 -0400, "Jones, Keven" <Keven.Jones_at_ncr.com>
wrote:
> Squid 3.0 is version. Can someone verify my squid.conf for me before I go
> to production with this?

Sure.

>
> Thanks to all in advance!!
>
> ------------------------------------------------
>
> http_port 80 accel defaultsite=img01.cprpt.com

Missing "vhost" option. Needed for multiple domain routing.

> cache_peer 172.19.23.91 parent 80 0 no-query originserver name=myAccel
>
> acl all src 0.0.0.0/0.0.0.0

Squid-3 defines "all" internally for you.
If you need to define it in the config then your version is too old and is
very buggy.

> acl our_sites dstdomain img01.cprpt.com
> http_access allow our_sites
> cache_peer_access myAccel allow our_sites
> cache_peer_access myAccel deny all
>
> cache_peer 172.19.23.92 parent 80 0 no-query originserver name=server_2
> acl sites_server_2 dstdomain img02.cprpt.com
> cache_peer_access server_2 allow sites_server_2

No "deny all" here? or did you intend this server to be a backup provider
of img01.cprpt.com as well?

>
> visible_hostname bv-ic01
>
> cache_dir ufs /data/spool/squid 100 16 256

100 MB of disk storage. You sure about that?

It's workable, but you may or may not want to alter it for better caching.
AUFS on linux and diskd on *BSD do better storage management than plain
ufs.

>
> cache_access_log /data/log/squid/access.log

Use instead:
access_log /data/log/squid/access.log

>
> cache_log /data/log/squid/cache.log
>
> cache_store_log /data/log/squid/store.log

Set the above to "none". The log is mostly a waste of disk IO time except
for debugging storage problems and some extremely detailed disk usage
analysis tools.

Amos
Received on Wed Sep 09 2009 - 23:36:47 MDT